基于激光测距成像和图像处理的输电线路防护技术

Protection Technology for Transmission Lines Based on Laser Range Imaging and Image Processing

针对目前输电线路防外力破坏自动化水平不高的现状,考虑激光扫描成像的特点,结合图像直线特征提取和空间转换,以及样本匹配理论,提出了一种在参数空间内的危险目标自动识别方法。通过脉冲飞行时间法来扫描获取目标各点的空间位置数据,并将原始数据映射为一幅灰度图像,将灰度图像进行图像处理,包括二值化、去噪、边缘检测,最终通过Hough变换提取出图像中的直线特征,在参数空间中,提出了判断图像间相似性的度量函数,并建立了目标图像与危险目标数据库的匹配函数,从而实现输电线下方危险目标的自动识别,防止输电线路遭受外力破坏。实验结果显示,能顺利检测到危险目标且不会发生误报,是一种可行的检测输电线路中危险目标的方法。

Under consideration of low level automatic protection of transmission lines against external damage, characteristics of laser scanimaging, linear image feature extraction and spatial switching, as well as sample matching theory, this paper presents an automatic recognition method for dangerous targets in parameter space, whereby spatial positional data of target points are obtained by scanning in the pulse flight time method and then the original data are mapped into a grayscale image. After image processing, including binarization, denoising and edge detection, we use Hough Transform to extract linear features of the image. For parameter space, we propose a metric function for judging similarity between images and establish a matching function between target image and dangerous target database thus realizing automatic recognition of dangerous targets under transmission lines to prevent external damage of these lines. Experimental results show that this approach can detect dangerous targets successfully without generating false alarm, and it is a feasible method for detecting dangerous targets on transmission lines.